Moving Target Tracking and Obstacle Avoidance for a Mobile Robot Using MPC

Abstract

Target tracking and obstacle avoidance are the most challenging tasks for autonomous non-holonomic mobile robots moving in unknown environments. In this paper a constrained model predictive controller is designed to track a moving target in an environment with static unknown obstacles. In the proposed controller both linear and angular velocities of the mobile robot are computed to control the next position and orientation of the robot. The design of tracking control problem is then formulated as a constraint quadratic programming (QP) problem. Solving the QP problem gives the vector of incremental control action, which will be used to calculate the optimal control input. The obstacle avoidance is also considered in designing the controller using a new switching algorithm and applying some changes in definition of the desired orientation. The proposed scheme guarantees the robot to track a moving target without collision with obstacles. The cost function has been defined as the Lyapunov function which proves stability of the system. The simulation results show a good performance for robot in tracking a moving target while avoiding some obstacles.